What This Means
This research examined what motivates older adults in the United States to be physically active, what barriers they face when using health and fitness apps, and how they feel about the potential use of artificial intelligence (AI) in those apps. Researchers conducted in-depth, hour-long interviews with 20 older adults and showed them images of AI-generated avatars to understand design preferences. The main finding was that older adults are primarily focused on maintaining their independence and ability to function day-to-day — not on becoming more fit or hitting performance goals. Many were already using wearable devices and smartphones to track their activity, and they engaged in both everyday tasks and structured exercise. While some found game-like features in apps motivating, others found them patronizing or stressful, highlighting how differently individuals respond to the same design choices.
A major theme was that current health apps present real accessibility challenges for older adults, including difficulty with touchscreens, cluttered screens, and complex navigation. Many preferred computers over smartphones for health-related tasks because of larger screens, keyboards, and perceived security. Despite these frustrations, participants were generally open to using AI in health apps — particularly if the AI could be personalized to their needs. Three different ways of thinking about AI emerged: some saw it as a simple tool to retrieve information, others imagined it as a coach giving guidance, and others wanted it to feel more like a companion. This variation suggests there is no one-size-fits-all design for AI in older adult health apps.
This research suggests that while older adults are willing to embrace AI-powered health technology, significant concerns remain about data privacy, whether AI would undermine their sense of independence, and whether AI recommendations can be trusted. Developers of health apps aimed at older adults may need to build in far more flexibility to accommodate individual preferences, and they may need to make AI systems more transparent and explainable to build user trust. The findings also highlight that accessibility improvements — simpler interfaces, larger text, and fewer complex gestures — remain a critical unmet need for this population.